Acute rejection is a frequent cause of early graft dysfunction in renal transplantation, and serum creatinine is the most important measure of clinical response. However, there is little information on serum creatinine during rejection episodes. In this study, the determinants that might influence the clinical response of serum creatinine during acute rejection were evaluated in 96 renal transplant recipients with 100 episodes of biopsy-proven rejection, by univariate and multivariate analysis. The factors assessed for their influence on serum creatinine in acute rejection included: presence and severity of vascular or cellular rejection on biopsy, time taken to institute therapy, HLA mismatch, and the use of OKT3. The presence of vascular rejection on biopsy (n = 28) was a major determinant of impaired reciprocal area under the curve (AUC) of serum creatinine (P < 0.01), and was correlated with HLA-A and -B mismatch (r = 0.21, P < 0.05). Acute rejection associated with HLA-DR mismatch resulted in a more rapid increase of serum creatinine, a higher maximal creatinine, and a greater AUC creatinine (all P < 0.05). Treatment of acute rejection with OKT3 had a beneficial effect on AUC creatinine (P < 0.05) when compared with intravenous corticosteroids, by multivariate analysis. However, vascular rejection and HLA-DR mismatch had no effect on serum creatinine 1 year after transplantation. The biopsy cellular rejection score had no effect on AUC creatinine, although there was a modest effect on gradient of creatinine prior to biopsy. A minor adverse effect of delay in therapy of acute rejection could be demonstrated in the methylprednisolone-treated subgroup. In summary, HLA-DR mismatch and the presence of vascular rejection were the most important predictors of the severity of rejection assessed by the response of serum creatinine to treatment. Appropriate antirejection therapy should be instituted promptly to optimize clinical response during acute rejection.